Mapping Language and Memory Functions in the Human Brain
National Institute of Neurological Disorders and Stroke, NIH
President Coates called the 2055th meeting to order at 8:29 p.m. on February 2, 1996. The Recording Secretary read the minutes of the 2052nd and of the 2054th meetings and they were approved. The President introduced Ms. Teresa Blaxton of the National Institute of Neurological Disorders and Stroke to discuss “Mapping Language and Memory Functions in the Human Brain”.
Ms. Blaxton began the presentation of her research on mapping functions of memory and language in the human brain with an illustration of early interest in mapping function onto structure, an 1840 newspaper advertisement for the phrenology services of “Silas Hubbard, MD” in Buffalo, NY. Modern research in this field began with the diagnosis, study and treatment of epilepsy. The seizures of epilepsy are now known to arise from uncontrolled electrical firing of neurons in the brain. The consequences of the disease for its victims are increased mortality, neuropsychological impairment, educational underachievement, underemployment and social isolation. 160,000 new cases are diagnosed each year in the United States. Of the approximately two million active cases, 300,000 experience uncontrolled seizures, and of those 50,000 are candidates for surgical intervention.
Electroencephalography had been helpful in diagnosing and assessing the treatment potential of epilepsy, but it provided little information useful in localizing the affected areas of the brain and assessing their function. Initially that important information could only be provided by direct electrical stimulation of the cerebrum of craniotomy patients, and that procedure is now performed almost exclusively during surgical treatment for epilepsy. Electrical stimulation is used to attempt to determine the function of the areas of the brain around the focus of neural discharges during epileptic seizures, to assess what tissues should and can be most safely removed. During these procedures it has been found that the seizure focus areas are frequently in parts of the brain apparently associated with speech and language.
Within the past 20 years several new noninvasive techniques have become available for revealing brain structures and mapping their function. One of these techniques is real time magnetic resonance imaging (MRI). It is most useful in mapping physical structures of the brain and blood flow through it. Ms. Blaxton illustrated the imaging potential of MRI with a reconstructed image of her own brain representing, as she hopes, a normal example. Unfortunately, in practice the signal to noise ratio for MRI is not yet sufficient for cognitive mapping studies; it can show where structures are but not what they are doing.
Another noninvasive technique useful in function mapping is positron emission tomography (PET). This technique uses injections of water labeled with oxygen-15 which decays by positron emission and then annihilation with photon pair production. The tomographic detection of the photon pairs is conducted on 15 planar sections of the whole brain with data integrated over 60 sec intervals. By indicating where recently injected radioactive water has been moving, this technique can reveal changes in regional cerebral blood flow and thus the differential activation of those brain regions. Because PET scans can be coordinated with the performance of cognitive tasks and monitored on a time scale of minutes, coordinated PET scans and cognitive tests have been helpful, especially in two regions of the brain thought to be involved with language, Broca's area in the anterior and Wernicke's area in the posterior. Coordinated PET scans and cognitive tests have also been helpful in identifying brain regions not traditionally thought to be critical to language.
One problem that must be addressed in designing experiments in this field is discriminating among the different cognitive tasks that may be involved. Another is developing baseline or null hypothesis tests to distinguish normal responses from background or non-responses, as well as from abnormal responses. Consider for example one common disability that has been the subject of some of this research — anomia, the inability to produce a name. Two aspects of the cognitive task of recalling a name can be recognized. One is the ability to recall a name by verbal association, that is by hearing spoken words or seeing written words, understanding their meaning and producing a spoken response. Another is the ability to recall a name by visual association, that is by viewing pictures, processing the image and producing a spoken response. When viewing words it is found by PET scanning that the bottom of the left hemisphere is activated. When viewing pictures it is the temporal lobe in Wernicke's area that is activated. Now, two types of memory may be involved in be able to recall words: conceptual processing analyzes semantic features, and perceptual processing analyzes physical features. Attempts to produce supposedly meaningless symbols or nonsense syllables to be used as null tests of perceptual processing is more difficult than might be presumed. Despite this, perceptual tasks have been created that do not stimulate the left meseal temporal region or the left lateral temporal but do stimulate the visual cortex. The nonverbal tests of perceptual tasks seem to cause stimulation in these regions on first exposure with decreased response on secondary exposures. This deactivation occurred in the auditory regions, but not apparently in the visual regions by these tests.
With these tests we can distinguish the anomia patients whose brains can recognize but can't write because of lexical inhibition, or can't say because of motor inhibition. In patients with semantic inhibition, the more description they provide the less likely they are to be able to recall the word — the flood of associations prevents the correct word from emerging. This approach has helped to advance our understanding of how memory processes are instantiated in the brain, distinguishing among neuroanatomical circuits engaged during some memory tasks but not others.
Ms. Blaxton kindly answered questions from the audience. The President thanked the speaker on behalf of the Society, announced the speaker for the next meeting, restated the parking policy, and adjourned the 2055th meeting at 9:43 p.m.
John S. Garavelli